Bitcoin's core developer, Gregory Maxwell, left Blockstream in January to focus on "deep-protocol work," and he and several other developers have now launched a proposal called "Erlay."
Save 75% of bandwidth
Erlay was announced on Tuesday by the University of British Columbia's Gleb Naumenko on the Bitcoin Core developer mailing list. Naumenko describes Erlay as follows:
- The node earns $20 a month? Lightning network in the eyes of the largest node operator
- Bitcoin Secret History: What did Satoshi Nakamoto and BM chat in Bitcointalk
- BTC returned to 10,000 points, the adjustment is over?
- A small step for Fidelity, a big step in the cryptocurrency market
- PayPal CEO responded for the first time “Libra exited the door” and revealed that individuals only hold bitcoin
- Can Bitcoin be $5,500 again? Bitcoin short-term price interpretation
“The basic concept is that information is sent directly through only a small number of connection points (only 8 output connections), and there is no need to publish every transaction to all peers. Periodically run on each connection between two-way reservation notification sets. Set up a coordination protocol to achieve further relaying… The result: we save half the bandwidth that the node needs to consume, allowing connections to be added almost freely, and better able to withstand timing attacks. If the outbound node count is increased to 32 Compared to the current protocol, Erlay will save 75% of the total bandwidth."
An important issue facing Bitcoin is how much bandwidth the node needs to consume in order to stay in sync. Bandwidth is one of the costs of running a full node. The Bitcoin blockchain currently has a capacity of more than 200 GB, which means that downloading a complete record can sometimes take days or even weeks.
By default, a node needs to connect to and receive information from approximately eight other nodes. Erlay supporters say that if this number is increased to 32, the bandwidth used by each node can be reduced by 75%.
Improve the security and efficiency of Bitcoin
Erlay introduces "diffusion" into Bitcoin, the opposite of the current model. The author of Erlay sees the current model as "version of flooding." “If a more efficient model is used, transactions and blocks can be spread across the network. The Erlay white paper states:
"The current version of the Bitcoin transaction relay protocol uses diffusion to propagate messages between nodes, which is a variant of random flooding. Flooding is a protocol in which each node advertises its reception to each of its peers. Each transaction is sent. Notifications can be sent via inbound and outbound links. With diffusion, peers can inject random delays before publishing received transactions. This reduces timing attacks and significantly reduces the likelihood of collisions. (When two nodes simultaneously publish the same transaction through a link between them)."
If the nodes are more interconnected, the specific attack vectors are also significantly reduced. "Timing attack" is a cryptographic concept that an attacker can exploit to break the cryptographic protocol by measuring the time required for the calculation. The Erlay protocol introduces random timing, so Bitcoin can be "better against timing attacks."
If the Erlay protocol is finally adopted, the latency and security of the Bitcoin network will improve. So far, the agreement has not been officially written as a bitcoin improvement, but we should be able to wait until this day. It is still an open discussion stage.